I Think I Can! (Do Physics)

Mmmm. It's been a refreshingly long time since I've had to read any arguments published by the New York Times about the innate inferiority of women in hard sciences. I'm looking at you, John Tierney.

(I did, however, read an entertaining rant by Natalie Angier about STEM, the annoying new acronym we're supposed to use instead of saying "hard sciences." "Aficionados pronounce STEM exactly as you'd imagine," she says, "like the plant part, like the cell type, like what you do to a tide and I wish I could do to this trend, but it's probably too late.")

While Tierney was polishing his bro card, researchers at the University of Colorado were conducting a pretty spiffy study about women in physics. For their subjects, they used an entire entry-level college physics class of 399 students (283 male and 116 female). As in any good psych study, the students were unaware of what was going on.

The study examined "stereotype threat," the phenomenon in which certain groups underperform when they're reminded of negative stereotypes about themselves. Specifically, the researchers wanted to know whether stereotype threat was bringing down the grades of the women in the class. Twice during the semester--once at the beginning, and again before the first midterm--all students completed a 15-minute writing assignment. Whether male or female, they were randomly assigned to write about either values that were important to them ("values affirmation") or values that were important to other people (control group). The values-affirmation exercise had been previously shown to help close the achievement gap for minority middle-school students. Neither the instructor nor the TAs knew which writing assignment the students had done.

At the end of the semester, the researchers looked at students' test scores, both on the usual class exams and a standardized physics test. They found that the values affirmation exercise didn't increase men's test scores--if anything, it lowered them a tiny bit. For the women, though, there was a clear jump from the control group to the values-affirmation group. In the control group, men outperformed women. After the values-affirmation exercise, the gap was gone.

Researchers also looked at the grade distribution in the class. While the values affirmation, again, hadn't affected men's grades, it clearly shifted a group of women out of the C range and into the B range. Interestingly, the percentage of women with A's didn't increase--it seems that the exercise was most helpful to the women who were struggling a bit to begin with. And there was one more piece of data: As part of an online course survey, the students had responded to the question, "According to my own personal beliefs, I expect men to generally do better in physics than women" on a five-point scale. It turned out that the women who agreed with this stereotype got the biggest boost from the values-affirmation exercise.

These students weren't insecure, math-averse middle-schoolers; they were college kids who were voluntarily taking physics because they wanted to major in science. It's surprising to see that stereotype threat could still have such a strong effect in this group. But it's encouraging that researchers are finding simple ways to counteract it. Or, unlike some columnists, acknowledging that there's a problem in the first place.